CavinHuang-comfyui-nodes-docs/docs/Unpack SDXL Tuple.md

Documentation

• Class name: Unpack SDXL Tuple
• Category: Efficiency Nodes/Misc
• Output node: False
• Repo Ref: https://github.com/comfyanonymous/ComfyUI

"Unpack SDXL Tuple"节点旨在将复杂的元组结构分解为其组成部分，专门用于高效处理与模型、剪辑和条件元素相关的数据。它促进了基础和精炼元素的分离和单独处理，提高了处理SDXL元组的模块化和灵活性。

Input types

Required

• sdxl_tuple
  • 需要解包的SDXL元组，包含基础和精炼阶段的组合模型、剪辑和条件信息。
  • Comfy dtype: SDXL_TUPLE
  • Python dtype: Tuple[torch.nn.Module, Any, str, str, torch.nn.Module, Any, str, str]

Output types

• BASE_MODEL
  • 从SDXL元组中提取的基础模型组件。
  • Comfy dtype: MODEL
  • Python dtype: torch.nn.Module
• BASE_CLIP
  • 从SDXL元组中提取的基础剪辑组件。
  • Comfy dtype: CLIP
  • Python dtype: Any
• BASE_CONDITIONING+
  • 从SDXL元组中提取的正面基础条件组件。
  • Comfy dtype: CONDITIONING
  • Python dtype: str
• BASE_CONDITIONING-
  • 从SDXL元组中提取的负面基础条件组件。
  • Comfy dtype: CONDITIONING
  • Python dtype: str
• REFINER_MODEL
  • 从SDXL元组中提取的精炼模型组件。
  • Comfy dtype: MODEL
  • Python dtype: torch.nn.Module
• REFINER_CLIP
  • 从SDXL元组中提取的精炼剪辑组件。
  • Comfy dtype: CLIP
  • Python dtype: Any
• REFINER_CONDITIONING+
  • 从SDXL元组中提取的正面精炼条件组件。
  • Comfy dtype: CONDITIONING
  • Python dtype: str
• REFINER_CONDITIONING-
  • 从SDXL元组中提取的负面精炼条件组件。
  • Comfy dtype: CONDITIONING
  • Python dtype: str

Usage tips

• Infra type: CPU
• Common nodes:
  • [Pack SDXL Tuple](../../efficiency-nodes-comfyui/Nodes/Pack SDXL Tuple.md)
  • TonemapNoiseWithRescaleCFG
  • FreeU_V2

Source code

class TSC_Unpack_SDXL_Tuple:

    @classmethod
    def INPUT_TYPES(cls):
        return {"required": {"sdxl_tuple": ("SDXL_TUPLE",)},}

    RETURN_TYPES = ("MODEL", "CLIP", "CONDITIONING","CONDITIONING", "MODEL", "CLIP", "CONDITIONING", "CONDITIONING",)
    RETURN_NAMES = ("BASE_MODEL", "BASE_CLIP", "BASE_CONDITIONING+", "BASE_CONDITIONING-",
                    "REFINER_MODEL", "REFINER_CLIP","REFINER_CONDITIONING+","REFINER_CONDITIONING-",)
    FUNCTION = "unpack_sdxl_tuple"
    CATEGORY = "Efficiency Nodes/Misc"

    def unpack_sdxl_tuple(self, sdxl_tuple):
        return (sdxl_tuple[0], sdxl_tuple[1],sdxl_tuple[2],sdxl_tuple[3],
                sdxl_tuple[4],sdxl_tuple[5],sdxl_tuple[6],sdxl_tuple[7],)